With the development of the Ethernet towards the direction of multi-service carrying, especially some services put forward increasingly high demand on reliability and real-time of the network, the Ethernet widely takes a form of a ring network, thereby improving the reliability of the network. Also, in the ring network protection solution, generally, it is required to achieve the fast protection switching below 50 ms. At present, the technology of fast protection switching is realized based on the G.8032 protocol of International Telecommunication Union (ITU-T).
Based on the above G.8032 protocol, an automatic protection switching protocol and mechanism are defined for the Ethernet layer of the ring topology Ethernet. Such network protection method is applicable for the ring topology Ethernet, and the implementation procedure thereof roughly includes: selecting a segment of a link as a ring protection link from the ring topology Ethernet, at least one of two adjacent nodes of the ring protection link blocking a port connected to the ring protection link when there is no failure in all the links of the Ethernet ring network to prevent protected data from passing through the ring protection link, in this way, only one communication path being between any two nodes of the Ethernet ring network, thus a closed loop of a communication path not occurring in the Ethernet ring network which avoids the closed loop and network storm; and when a link of the Ethernet ring network fails, if the failed link is not the ring protection link, blocking the adjacent port of the ring protection link, and unblocking the blocked port in the ring protection link so that the protected data can pass through the ring protection link, and a new communication path is generated between any two nodes of the ring protection link, thus ensuring re-connection of the communication path and improving the reliability of the network.
During the switching process of the Ethernet ring network, the communication path of data traffic needs to be changed, and since address table information recorded by each node is not suitable after the communication path of data traffic has been changed, each node needs to update the address table, and an specific action is to delete address entries in the ports on ring. In the G.8032 protocol, when a node sends a protocol frame, the protocol frame includes its own node number and a blocked port index number, and the other nodes will extract the node number and the blocked port index number in the protocol frame after receiving the protocol frame and compare them with a node number and a blocked port index number originally stored in refreshing record information corresponding to a receiving end. If they are different, then the refreshing record information corresponding to the receiving end will be updated as the extracted node number and blocked port index number, and if the node number and the blocked port index number in the protocol frame are different from a node number and a blocked port index number of refreshing record information corresponding to an opposite end of the receiving end, then this node may further need to refresh the address table.
Although the above method for refreshing the address table of the Ethernet ring network guarantees that wrong address entries can be quickly deleted during the switching of the Ethernet, a problem of repeating refreshing during switching is further brought about. For example, in a scene that a link is switched back after being restored, the node to which the ring protection link belongs may repeat refreshing the address table, because when the node to which the ring protection link belongs is switched back, it will first block an adjacent port of the ring protection link, and then send a ring protection link blocked protocol frame to notify other nodes to be switched back, and subsequently, the node to which the ring protection link belongs will refresh the address table. However, when the node to which the ring protection link belongs then first receives the ring protection link blocked protocol frame sent by itself, as the node number and the blocked port index number of the refreshing record information of the original port are different from information included in the received protocol frame, the node to which the ring protection link belongs repeats refreshing the address table once again, which in fact is unnecessary, and will lead to broadcasting the traffic again after being stable. The problem that the node to which the ring protection link belongs repeats refreshing when the Ethernet is switched back results from the blocked port information of the node to which the ring protection link belongs in the Ethernet not being able to be updated in time. At present, there is still no scheme for the problem that the blocked port information cannot be updated in time.